SUMO adaptive traffic signal control - DQN, DDPG, Webster's, Max-pressure, Self-Organizing Traffic Lights
Technical details available at An Open-Source Framework for Adaptive Traffic Signal Control
Install SUMO traffic microsimulator by following instructions here (v1.2).
Using Python 3, create a virtual environment and then install dependancies with:
pip install -r requirements.txt
First train reinforcement learning controllers:
./train_dqn.sh
./train_ddpg.sh
Then execute simulations to generate performance results for all controllers:
./gen_results.sh
Visualize results with:
python graph_results.py
Search for optimal hyperparameters for each controller:
./hp_optimization
Warning, search for reinforcement learning can require significant compute time!
Visualize hyperparameters with:
python graph_results.py -type hp